Aspects of Respiratory Muscle Fatigue in a Mountain Ultramarathon Race

THOMAS U. WU¨ THRICH1, JULIA MARTY1, HUGO KERHERVE2, GUILLAUME Y. MILLET3,4, SAMUEL VERGES5,6, and CHRISTINA M. SPENGLER1,7 1Exercise Physiology Lab, Institute of Human Movement Sciences and Sport, ETH Zurich, SWITZERLAND; 2School of Health and Exercise Science, University of the Sunshine Coast, Queensland, AUSTRALIA; 3Exercise Physiology Laboratory, University of Lyon, Saint-E´ tienne, FRANCE; 4Human Performance Laboratory, Faculty of Kinesiology, University of Calgary, Calgary, CANADA; 5HP2 Laboratory, University Grenoble Alpes, Grenoble, FRANCE; 6U1042, INSERM, Grenoble, FRANCE; and 7Zurich Center for Integrative Human Physiology, University of Zurich, Zurich, SWITZERLAND

ABSTRACT WU¨ THRICH, T. U., J. MARTY, H. KERHERVE, G. Y. MILLET, S. VERGES, and C. M. SPENGLER. Aspects of Respiratory Muscle AI SCIENCES BASIC Fatigue in a Mountain Ultramarathon Race. Med. Sci. Sports Exerc., Vol. 47, No. 3, pp. 519–527, 2015. Purpose: Ultramarathon running offers a unique possibility to investigate the mechanisms contributing to the limitation of endurance performance. Investigations of locomotor muscle fatigue show that central fatigue is a major contributor to the loss of strength in the lower limbs after an ultramarathon. In addition, respiratory muscle fatigue is known to limit exercise performance, but only limited data are available on changes in respiratory muscle function after ultramarathon running and it is not known whether the observed impairment is caused by peripheral and/or central fatigue. Methods: In 22 experienced ultra-trail runners, we assessed respiratory muscle strength, i.e., maximal voluntary inspiratory and expiratory pressures, mouth twitch pressure (n = 16), and voluntary activation (n = 16) using cervical magnetic stimulation, lung function, and maximal voluntary ventilation before and after a 110-km mountain ultramarathon with 5862 m of positive elevation gain. Results: Both maximal voluntary inspiratory (j16% T 13%) and expiratory pressures (j21% T 14%) were significantly reduced after the race. Fatigue of inspiratory muscles likely resulted from substantial peripheral fatigue (reduction in mouth twitch pressure, j19% T 15%; P G 0.01), as voluntary activation (j3% T 6%, P = 0.09) only tended to be decreased, suggesting negligible or only mild levels of central fatigue. Forced vital capacity remained unchanged, whereas forced expiratory volume in 1 s, peak inspiratory and expiratory flow rates, and maximal voluntary ventilation were significantly reduced (P G 0.05). Conclusions: Ultraendurance running reduces respiratory muscle strength for inspiratory muscles shown to result from significant peripheral muscle fatigue with only little contribution of central fatigue. This is in contrast to findings in locomotor muscles. Whether this difference between muscle groups results from inherent neuromuscular differences, their specific pattern of loading or other reasons remain to be clarified. Key Words: MOUTH TWITCH PRESSURE, ULTRAENDURANCE, RUNNING, POSTURE, CERVICAL MAGNETIC STIMULATION

ince the 1970s, ultraendurance events have become respiratory muscle strength (19,39) after ultraendurance run- increasingly popular in the Western world (15). Such ning ranging from 80 to 154 km, but results are conflicting. Sevents provide the unique possibility to study physi- Warren et al. (39) reported no change in both forced vital ology when the human body is pushed toward its limits (25). capacity (FVC) and forced expiratory volume in 1 s (FEV1) Several studies have investigated different phenomena after a 24-h race, and Mahler and Loke (22) reported sig- associated with ultraendurance-type competitions such as the nificant reductions in FVC (j12%) and FEV1 (j10%) after occurrence of neuromuscular fatigue of the locomotor system an 81-km ultramarathon competition; neither of them found (e.g., (27)). However, the pulmonary system and particularly significant changes in maximal inspiratory pressure (MIP) the respiratory muscles have not yet been conclusively in- and maximal expiratory pressure (MEP) after these races. vestigated in this type of competition. So far, only three This may have two reasons, as follows: Warren et al. (39) studies assessed changes in either lung function (22,39) or reported large interindividual variability in the reduction of MIP and MEP pressure (0% to j37%) in a small number of subjects (n = 10), whereas Ker and Schultz (19) performed the after-race measurements not immediately but 3 d after Address for correspondence: Christina M. Spengler, Ph.D., M.D., Exercise Physiology Lab, ETH Zurich, Winterthurerstrasse 190, 8057 Zurich, the race. Interestingly, despite three days of recovery, these Switzerland; E-mail: [email protected]. authors measured 27% reduction in time to exhaustion in a Submitted for publication April 2014. respiratory muscle endurance test. After marathon running Accepted for publication July 2014. (i.e., 42.2 km), MIP was significantly decreased by 15%– 0195-9131/15/4703-0519/0 25% (6,21,31) whereas MEP was found to be decreased by

MEDICINE & SCIENCE IN SPORTS & EXERCISEÒ 28% (21). Given the task-specific nature of fatigue (9), the Copyright Ó 2014 by the American College of Sports Medicine extent to which respiratory muscle fatigue occurs during DOI: 10.1249/MSS.0000000000000449 ultramarathon running remains to be clarified.

519

Copyright © 2015 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited. Neuromuscular fatigue is defined as ‘‘I aconditionin different reasons such as excessive coughing, dizziness, or which there is a loss in the capacity for developing force general uneasiness. The remaining 23 participants (eight and/or velocity of a muscle, resulting from muscle activity women) had normal lung function and respiratory muscle under load and which is reversible by rest’’ (28). Importantly, strength (Table 1). All participants were experienced ultra- fatigue can occur at different levels, from the brain to the marathon runners, which was a requirement for participation skeletal muscles (11). None of the existing studies in both in the competition supporting this study (North-FaceÒ Ultra- ultraendurance events and marathon running performed Trail du Mont BlancÒ 2012). measurements to distinguish between central and peripheral All participants were fully informed about the procedures mechanisms of respiratory muscle fatigue, i.e., measurements and risks involved in this study and that they could withdraw of muscle contractility during nerve or muscle stimulation. from the study at will at any time. Each participant gave Interestingly, assessments in exercise challenges of shorter his/her written informed consent. The study was approved duration showed that peripheral changes in respiratory muscle by the local ethics committee (Comite´ de Protection des contractility are usually substantially greater than changes Personnes Sud-Est 1, France) and was performed according observed in voluntary maneuvers such as MIP and MEP (17), to the Declaration of Helsinki. which are, in fact, similar to observations in locomotor muscles (26). Given that substantial levels of central fatigue have been Experimental Overview reported for locomotor muscles after ultraendurance running (27), some of the reduction in maximal voluntary respiratory Each participant visited the laboratory three times. During muscle strength assessed with voluntary maneuvers may also the first visit, which was scheduled 3–4 wk before the be attributed to central fatigue, i.e., reduction in maximal drive competition, participants performed an incremental test and to the contracting muscles. In fact, peripheral respiratory mus- were introduced to the experimental procedures (for details, cle fatigue has traditionally been associated with high levels see the following section). The second visit (before race) of ventilation observed during intense exercise (17). Additional took place within the 3 d before the race and included loads are placed upon respiratory muscles in ultramarathon baseline measurements of respiratory muscle strength and

BASIC SCIENCES running apart from ventilation, which could impair respiratory lung function (for details, see the following section). The muscle contractility as well, one of them being the postural third visit (after race) took place 1.4 T 0.4 h after completion demand (12). Moreover, it has been suggested that an in- of the race, and procedures were identical to those of the pre- crease in inflammatory mediators (e.g., tumor necrosis factor race assessment. The 2012 edition of the competition >) may also alter muscle contractility (40) and/or the capacity supporting this study was shortened to 110 km (instead of to produce maximal muscle force (27). Thus, the contribution 167 km) because of bad weather conditions (rain and snow), of peripheral and central mechanisms to respiratory muscle and it included a total positive elevation gain of 5862 m. The fatigue after an ultraendurance run remains to be clarified. maximal temperature during the race was 12-C in Chamonix To determine the development of respiratory muscle fa- (1459 m above sea level) and dropped below 0-C at altitudes tigue, both maximal inspiratory and expiratory muscle above 1800 m. strength were assessed by volitional MIP and MEP maneu- First visit. After medical examination, participants per- vers before and after a mountain ultramarathon (110 km). formed an exhaustive incremental test on a treadmill For inspiratory muscles, we also assessed mouth twitch (EF1800; HEF Tecmachine, Andre´zieux-Bouthe´on, France), pressure (Pm,tw) in response to cervical magnetic stimulation with a 10% grade to determine maximal oxygen consump- to better understand the origin of potential loss in volitional tion. The protocol started at 4–6 kmIhj1 depending on run- muscle strength after the race. In addition, we determined ning ability of the participant and was then increased by changes in lung function to account for potential changes in 1kmIhj1 every 2.5 min. At the end of each stage, partici- lung volume that would affect measures of respiratory pants were stopped for 30 s for blood sampling to assess muscle strength and to add to the sparse data available on the blood lactate concentration. Ventilation and gas exchange effect of ultramarathon running on pulmonary function. We were obtained breath by breath using a metabolic cart hypothesized that inspiratory muscle fatigue would develop (Ergocard; Medisoft, Sorinnes, Belgium) and were averaged during the ultramarathon because of both central and pe- over 30-s intervals. Maximal oxygen consumption was de- ripheral mechanisms, measured by reductions in Pm,tw,in fined as the fulfillment of at least three of the four following volitional respiratory muscle strength and activation and in criteria during the last 30 s before reaching exhaustion: 1) no flow-dependent variables of lung function measures. further increase in oxygen uptake with increasing workload,

TABLE 1. Subjects characteristics. METHODS Age (yr) 42.9 T 9.4 MIP (%pred) 125 T 26 Height (m) 1.72 T 9.0 MEP (%pred) 161 T 28 Participants and Ethical Statement Body mass (kg) 67.0 T 9.4 FVC (%pred) 92 T 9 j1 j1 V˙ O2max (mLImin Ikg ) 57.2 T 6.1 FEV1 (%pred) 103 T 11 Thirty-five runners were initially recruited for the present MVV (%pred) 114 T 32 PEF (%pred) 104 T 12 study. Twelve runners either did not finish the race (n =6) Values are mean T SD. or could not perform post-race measurements (n = 6) for %pred, percent of predicted value; V˙ O2max, maximal oxygen consumption.

520 Official Journal of the American College of Sports Medicine http://www.acsm-msse.org

Copyright © 2015 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited. 2) RER 9 1.1, 3) achievement of the age-predicted HRmax output. Special care was taken to ensure that inspiratory within 10 bpm, and 4) blood lactate concentration 9 8 mM. muscles were in a fully potentiated state. Potentiated muscle All subjects met this requirement. twitches were evoked, as they are known to be more sensi- Second (before) and third (after) visit. The second tive to changes occurring within a fatigued muscle (20). and third visits were conducted at a laboratory near the Pm,tw responses were assessed after 3–4 submaximal and finishing area of the race. On the second visit, all partici- three maximal inspiratory efforts at FRC against a closed pants were first thoroughly familiarized with the assessment airway, maximal efforts lasting for 5 s. After the third and of respiratory muscle strength and lung function. Thereafter, sixth stimulations, another 5-s maximal inspiratory effort participants performed all measurements in the following followed to maintain the potentiated state. Moreover, during order: cervical magnetic stimulation protocol, assessment of the third, fourth, and fifth maximal inspiratory effort, a sin- MIP and MEP, and lung function (details are in the fol- gle stimulation was superimposed onto the maximal inspi- lowing section). On the third visit, measurements were re- ratory effort. The ratio between the superimposed twitch peated in the same order. amplitude and that of the following Pm,tw was used to cal- culate the maximal voluntary activation (VA) as follows: Assessment of Respiratory Muscle Contractility superimposed twitch amplitude Volitional respiratory muscle strength. MIP (from VAð%Þ¼1jA 100 amplitude of Pm;tw residual volume) and MEP (from total lung capacity) were AI SCIENCES BASIC measured using a handheld mouth pressure meter (Micro A correction term A was included to correct for RPM; Micro Medical Ltd., Rochester, United Kingdom) superimposed stimulation, which was not delivered at the to obtain volitional maximal inspiratory and expiratory highest volitionally achieved Pm. strength. Participants performed a minimum of three ma- volitional P just before superimposed twitch A ¼ m neuvers or until values did not improve anymore. The larger highest volitional Pm during maneuver of the two maximal values that did not differ by more than 5% was selected as the maximum. Percent predicted values To ensure supramaximal stimulation of the phrenic were calculated according to reference values (41). nerves, three additional twitches were performed at 98% and Nonvolitional inspiratory muscle strength. Phrenic 94% of the stimulator output (each group of three stimula- nerves were stimulated on the back of the flexed neck be- tions was preceded by a 5-s maximal inspiratory effort). tween the sixth and seventh cervical vertebra to induce mus- Because of time restrictions, this was only done in the fa- cle twitches. Cervical magnetic stimulation was performed tigued state after the race (see Discussion). A near plateau in using a MagStim 200 stimulator (max. 2 T, 1-ms rectangular Pm,tw was observed in every participant, and no statistically pulses; MagStim, Whitland, United Kingdom) equipped with a significant difference was found between 98% and 100% of 90-mm circular coil. During all stimulations, participants were the stimulator output (P = 0.57), thereby indicating maximal comfortablyseatedonachair,breathing through a mouthpiece recruitment of all motor nerves (see Discussion). This is connected to a heated pneumotachograph (MLT3813H-V; similar to previous reports using the same protocol to test for Hans Rudolph Inc., Shawnee, KS) that was attached to a dif- supramaximal stimulation in our laboratory (37). ferential pressure transducer (FE141; ADInstruments, Bella For Pm,tw, the average amplitude (baseline to peak) of 3–9 Vista, Australia) to measure respiratory flow. The measure- potentiated twitches as well as contraction time (i.e., dura- ment system also included a shutter for short airway occlusions tion from stimulation to the peak of Pm,tw) and the maximal during stimulations (Zan, Oberthulba, Germany). Before each rate of pressure development (i.e., the largest incline of the stimulation, participants were instructed to passively expire to Pm,tw response) were calculated. A twitch was rejected functional residual capacity (FRC). As soon as the point of post hoc if the end-expiratory lung volume deviated from zero flow was reached (i.e., FRC), the shutter closed and sub- FRC by more than T5% of vital capacity. jects were instructed to gently inspire until mouth pressure (P ) m Assessment of Lung Function reached j5cmH2O, which triggered the release of the stimulation (18). FRC was obtained during a 2-min resting period Lung function was assessed according to current American before the start of the cervical magnetic stimulation protocol. Thoracic Society/European Respiratory Society guidelines End-expiratory lung volume at FRC was placed within vital (2) using a metabolic cart with a calibrated volume sensor capacity maneuvers to compare between measurements before (Quark b2; COSMED, Rome, Italy). Percent predicted and after the race. The pressure response of each muscle twitch values of lung function were calculated according to refer- was obtained at the mouth (i.e., Pm,tw) using a calibrated ence values (30). pressure transducer (DPT-100; Utah Medical Products Ltd., Data Analysis and Statistics Athlone, Republic of Ireland). Figure 1 shows pressure, flow, and volume traces for a single stimulation in a representative For before-and-after race comparison, paired t-tests were subject before and after the race. conducted after analyzing for normal distribution (Shapiro–Wilk The cervical magnetic stimulation protocol consisted of test). For data without normal distribution, a Wilcoxon signed nine potentiated Pm,tw responses at 100% of the stimulator rank test was applied. Statistical analyses were performed with

RESPIRATORY MUSCLE FUNCTION IN ULTRAMARATHON Medicine & Science in Sports & Exercised 521

FIGURE 1—Raw data recordings of a representative subject showing pressure, flow, and volume during a single cervical magnetic stimulation obtained before and after the race. VC, vital capacity.

SPSS Statistics 19 (IBM Company, New York, NY). All data participants Pm,tw could not be assessed appropriately after are shown as mean T SD. The level of significance was set at the race because of technical problems; thus, the number P G 0.05 for all statistical comparisons. of participants varies slightly among parameters. Both MIP and MEP were significantly reduced (P G 0.01, n = 22) after the race, along with significant decrease in Pm,tw (P G 0.01, RESULTS n = 16) and a tendency for VA to be reduced (P = 0.09, Ultramarathon performance. Approximately 2500 n = 16). Twitch contraction time before (96 T 13 ms) and runners participated, and 85% successfully finished the race after (94 T 8 ms) remained unchanged (P = 0.35, n = 16), (83% of the study participants finished). The average time to whereas maximal rate of pressure development after the race j1 complete the race for participants in the present study was (578 T 195 cm H2OIs ) was significantly lower than that j1 20.1 T 3.4 h (range, 13.8–25.8 h), resulting in an average before the race (702 T 185 cm H2OIs , P G 0.01, n = 16). velocity of 5.3 T 0.9 kmIhj1. All participants were able End-expiratory lung volume before cervical magnetic to cover the final kilometer of the race (over a flat portion stimulation did not differ between measurements before and of the track) at a much higher pace (average, 12.1 T after the race, with an average change of 0.017 T 0.075 mL 1.7 kmIhj1; range, 9.8–16.8 kmIhj1) than their respective (P = 0.39), corresponding to 0.5% T 1.6% of FRC. The average velocity. average within-session coefficient of variation (CV) for Respiratory muscle strength. Average and individ- Pm,tw was 7.1% T 3.6%. ual changes in MIP, MEP, Pm,tw, and VA are displayed in Lung function. Absolute values and relative changes in Figure 2. One participant did not perform any respiratory lung function parameters and maximal voluntary ventilation strength measurement, three participants only performed (MVV) are given in Table 2. One participant could not volitional strength measurements (MIP and MEP), and three perform lung function measurements after the race because

522 Official Journal of the American College of Sports Medicine http://www.acsm-msse.org

FIGURE 2—Individual changes in Pm,tw (n = 16), volitional MIP (n = 22), MEP (n = 22), and VA (n = 16) from before (set at 100%) to after the race. Open circles represent individual data, whereas closed circles depict average values. *Significantly different from values before the race, P G 0.01. of airway irritation. Apart from FVC, all lung function pa- thereby offering new insights into neuromuscular conse- rameters were significantly reduced after the race (P G 0.05, quences for the respiratory system in such exercise. n = 22). Reductions in variables of respiratory muscle strength and reductions in peak inspiratory flow (PIF), peak Neuromuscular Fatigue Induced by an expiratory flow (PEF), and MVV were not correlated. Ultramarathon Mountain Race

DISCUSSION The significant reduction in voluntary respiratory muscle strength (assessed by MIP and MEP) found in the present The main purpose of the present study was to investigate study is in contrast to findings obtained elsewhere, reporting whether inspiratory and/or expiratory muscle fatigue would no systematic change in MIP and MEP after ultraendurance develop during an ultraendurance running competition and, competitions (19,39). This discrepancy may be explained by if so, to what extent peripheral versus central mechanisms either the prolonged time between the end of the competition contribute to the loss of volitional strength. We found 1) that and the measurements or by an insufficient sample size in significant inspiratory and expiratory muscle fatigue devel- previous studies. In fact, Warren et al. (39) reported nonsig- oped as indicated by substantial reductions in MIP (j19%) nificant MIP reduction, although seven of 10 participants had and MEP (j21%), 2) that peripheral inspiratory muscle fa- reduced values after the race. Similarly, they reported large tigue, i.e., loss in muscle contractility, contributed signifi- variability in the MEP reduction after the race, which was cantly to the decrease in volitional inspiratory muscle likely the cause for the decrease in MEP not reaching the strength, as shown by reduction in Pm,tw (j19%) whereas level of statistical significance. When comparing the present voluntary muscle activation only tended to decrease (j3%), findings with those found after marathon running, similar and 3) that all flow-dependent variables of lung function reductions in maximal voluntary respiratory muscle strength assessments were significantly reduced. The present study were observed (i.e., 15%–25%), thereby confirming the provides, for the first time, nonvolitional measurements of presence of respiratory muscle fatigue after endurance com- inspiratory muscle strength (using cervical magnetic stimu- petitions of low-to-moderate intensities (6,21,31). Interest- lation) before and after an extreme endurance running event, ingly, both MIP and MEP reductions seem to be substantially

TABLE 2. Lung function parameters. j1 j1 j1 j1 FVC (L) FEV1 (L) PEF (LIs ) MEF50% (LIs ) PIF (LIs ) MVV (LImin ) Before race 4.22 T 0.88 3.78 T 0.75 8.60 T 1.81 4.96 T 1.26 6.33 T 2.09 149 T 32 After race 4.11 T 0.84 3.49 T 0.78* 7.41 T 1.98* 4.26 T 1.42* 5.20 T 1.72* 132 T 34* Percent reduction j1.8 T 11.3 j7.3 T 11.6 j13.8 T 13.4 j14.0 T 17.8 j16.8 T 12.0 j16.8 T 12.0 Values are mean T SD (n = 22). *Significantly different from values before the race, P G 0.05). MEF50%, mid expiratory flow at 50% of FVC.

RESPIRATORY MUSCLE FUNCTION IN ULTRAMARATHON Medicine & Science in Sports & Exercised 523

Copyright © 2015 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited. larger in the present study than reductions seen for short- of fatigue found in respiratory and locomotor muscles after duration high-intensity exercise that are reported to be about ultramarathon competitions. These differences could be re- 10% or less (e.g., (23). These results suggest that similar to lated to the load imposed on respiratory and locomotor locomotor muscle fatigue, the relation between the loss of muscles or differences regarding their intrinsic properties. volitional respiratory muscle strength and the increase in ex- This, and the level of central fatigue contributing to the post- ercise duration follows an asymptotic relation reaching a race loss in MEP, remains to be investigated. As opposed to plateau for respiratory muscle fatigue during running exer- central fatigue, a significant amount of peripheral inspiratory cises with the length of marathon or longer races (26). muscle fatigue was observed, as highlighted by 19% re- To elucidate the origin of global inspiratory muscle fa- duction in Pm,tw. In 10 of 16 participants (i.e., 63%), a rel- tigue, cervical magnetic stimulation was performed before evant amount of fatigue was detected, which was defined as the race and within 1 h after completion. Surprisingly, only Pm,tw reduction exceeding 15%, i.e., 2 times the average minor changes in VA were found after the race (approxi- within-session CV. This corresponds well to the prevalence mately 3%), which contradicts our original hypothesis. This of inspiratory muscle fatigue (approximately 72%) assessed suggests that the contribution of central fatigue to the ob- by artificially induced twitch pressures after high-intensity served reduction in MIP was small. Bellemare and Bigland- exercise in our and other laboratories (17,34). Whether ex- Ritchie (3) were the first to introduce VA to estimate piratory muscle fatigue also occurs at or distal to the neu- central components of diaphragm fatigue, and de Bruin et al. romuscular junction remains to be investigated because we (8) confirmed that this technique may also be valid when have not yet established a noninvasive measure of peripheral applied to superimposed Pm,tw, similar to those applied in expiratory muscle fatigue, i.e., without using gastric balloon the present study. Bellemare and Bigland-Ritchie (4) also catheters. Therefore, we could not distinguish between pe- showed that central fatigue can occur in the diaphragm, ripheral and central expiratory fatigue in the present study. as indicated by substantial reduction in VA after resistive However, because both the muscles of inspiration and ex- breathing to exhaustion. However, they only investigated piration share similar tasks during exercise, e.g., breathing five individuals in that study and resistive breathing does not and trunk stabilization, one could speculate that similar

BASIC SCIENCES represent the load that inspiratory muscles are exposed to mechanisms were involved in the development of both in- during exercise hyperpnoea. Therefore, direct comparison spiratory and expiratory muscle fatigue. This, however, re- with our results seems difficult. To the best of our knowl- mains to be tested. edge, no study has yet distinguished between central and Peripheral fatigue may either originate from reduced ex- peripheral components of inspiratory muscle fatigue after citability (i.e., decrease in action potential propagation and/or competitive endurance events. Using the superimposed transmission) or changes within the excitation–contraction twitch technique, the present data suggest that even after an coupling mechanism. Although not assessed in this study, ultraendurance race, respiratory muscle activation is well the former is unlikely because altered M-wave characteris- preserved. This is in contrast to previous results indicat- tics have not been reported for fatigued inspiratory muscles ing substantial central fatigue in locomotor muscles (27,35). in our or in other laboratories (17,37). Thus, it seems likely When comparing mechanisms of respiratory muscle fatigue that peripheral fatigue measured approximately 1 h after in the present study with locomotor muscle fatigue in these the race was mostly due to impaired excitation–contraction latter studies, a potential difference in recovery time needs to coupling mechanisms (1). These cellular processes can be be considered. In fact, Millet et al. (27) obtained knee ex- influenced by different metabolic factors, of which reduced tensor muscle strength approximately 20 min after the race muscle glycogen seems to be most relevant in light of the whereas Temesi et al. (35) reassessed their subjects ap- character of the race and the elapsed time until after-race proximately 1 h after crossing the finishing line. Despite this measurements could be performed (29). This is based on difference in the period of potential recovery, both authors animal studies showing that glycogen content in the dia- reported an average 19% reduction in VA, which may sug- phragm was substantially reduced after endurance-type ex- gest that recovery of central fatigue is rather slow. In the ercise, which persisted for several hours (13). The cause of present study, post-race measurements were performed ap- these changes in the intramuscular milieu is likely to be proximately 1.4 h after the end of the competition on aver- multifactorial. First, the level of ventilation itself needs to be age, with no significant reduction in VA. Although this adds considered. In fact, the average running velocity during the 24 min of recovery compared with that in the study of race was 5.3 kmIhj1, suggesting rather modest levels of Temesi et al. (35), it seems reasonable to suggest that central ventilation. However, short intervals of higher metabolic fatigue of respiratory muscles was likely not much greater demand such as during steep inclines or during the final immediately after the race given that the recovery rate of sprint possibly increased ventilation to levels known to af- respiratory muscles is similar to that of locomotor muscles. fect respiratory contractility (e.g., (17)). In a population of This is further supported by the fact that, in the present trained runners, for instance, walking at 5 kmIhj1 with 16% study, the reduction in VA did not correlate with the dura- grade was accompanied by a ventilation of approximately tion of recovery (data not shown). Therefore, other expla- 60 LIminj1 in normoxia, corresponding to the level of nations should be considered for the different mechanisms ventilation during running at approximately 11 kmIhj1 (5).

524 Official Journal of the American College of Sports Medicine http://www.acsm-msse.org

Copyright © 2015 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited. On average, participants ran the final kilometer at 12 kmIhj1 have been present, as Miles et al. (24) showed that despite (range, 9.8–16.8 kmIhj1), and in several stages of the race, unchanged FVC, a mild transient subclinical edema was they were exposed to moderate hypoxic conditions, which likely to be present after a marathon race, indicated by can both contribute to higher levels of ventilation and the substantial reduction in pulmonary diffusing capacity that development of respiratory muscle fatigue (36). Yet, we did was, unfortunately, not assessed here. Thus, further investi- not find any correlation of the velocity during the final ki- gations including techniques independent from the partic- lometer and the reduction in Pm,tw. ipant’s effort are required to gain deeper insights into the An alternative explanation apart from ventilation may be potential presence of peripheral airway constriction and/or found in the increased postural demand placed upon respi- peribronchial edema in ultramarathon running. ratory muscles during uphill and downhill running on un- even surfaces, such as that encountered during a mountain Implications of Respiratory Muscle Fatigue for ultramarathon. Several lines of evidence exist suggesting (Mountain) Ultramarathon Races that the diaphragm—for instance—is strongly involved in The observed reductions in P induced by cervical the stabilization of the trunk (12). Lastly, inflammatory m,tw magnetic stimulation are likely to represent global inspira- agents could also affect muscle contractility and activation tory muscle fatigue rather than isolated diaphragm fatigue (40). Ultramarathons were shown to trigger a very pro- (32). In turn, global inspiratory muscle fatigue in healthy nounced systemic inflammatory response, which could then individuals has recently been shown to affect postural con- SCIENCES BASIC have detrimental effects on muscle contractility (27). trol by inducing a shift to a more rigid control strategy in- Changes in Lung Function after volving mainly the ankles, which is associated with decreased Ultramarathon Running stability (16). Taking into account the massive reduction in calf muscle and quadriceps muscle strength after a similar Only two studies have assessed the changes in lung func- ultramarathon race (27,38), it is tempting to speculate that tion immediately after ultramarathon competitions, and re- with the occurrence of both respiratory and leg muscle fa- sults are conflicting. FVC was not altered in the present study, tigue, runners might be exposed to elevated risk of fall and which is in line with the study of Warren et al. (39), whereas injury especially when running on challenging undergrounds Mahler and Loke (22) found a 12% reduction after an 80- to as present in mountain marathon races. This hypothesis, how- 100-km race of markedly shorter duration (i.e., approximately ever, needs further investigation. 6–10 h) than that in the present study and, thus, higher exercise intensity. In the present study, FEV (j7%), PEF 1 Methodological Considerations (j14%), mid-expiratory flow at 50% of FVC (j14%), PIF (j17%), and MVV (j14%) all significantly decreased, Variation in respiratory muscle strength and lung which is mostly in line with previous findings (22,39). function. The reproducibility of Pm,tw measurements is a Because of the nature of lung function assessment using crucial element to the outcome of this study. Because of volitional maneuvers, fatigue per se could have contributed the study design, only within-session CV (on average, 7.1%) to reductions in flow-dependent measures. However, none but not between-day CV could be assessed. However, un- of the changes in MIP, MEP, and Pm,tw correlated with published data from our laboratory estimated between-day changes in flow-dependent variables of lung function. CV for Pm,tw to be approximately 7.0%. Therefore, we are However, the missing direct link between muscle fatigue confident that the chosen 15% threshold (i.e., twofold CV) and lung function reductions in this study should be to detect a physiologically relevant reduction in Pm,tw in- interpreted with caution because static measures such as dicates true alterations in muscle contractility. MIP and MEP (placed at one end of the muscle force– A factor contributing to this variability could reside in velocity curve) do not include changes in the potential to changes of lung volume at the time of stimulation, as it was produce high shortening velocities as present during all shown that lung volume can substantially affect twitch flow-dependent maneuvers. The present data show, in fact, pressure amplitude (14). However, end-expiratory lung that the maximal rate of force development of the inspiratory volume was tightly controlled in the present study via con- muscles was significantly reduced. Thus, part of the re- tinuous monitoring of the volume channel, thereby ensuring ductions in dynamic lung function variables may well be that the same lung volume was reached before each stimu- associated with muscular fatigue per se. lation. If lung volume before stimulation deviated by more Additionally or alternatively, impairment of lung function than T5% from FRC, data were excluded post hoc. From may also be influenced by other factors such as peripheral before to after the race, lung volume before stimulation only airway constriction and/or peribronchial edema. However, changed by 0.02 L (range, j0.09 to 0.14 L), corresponding only reductions in FEV1 greater than 10% are indicative of to 0.5% (range, j1.5% to 3.8%) on average. Using the peripheral airway constriction (7), which was observed in equation provided by Hamnegård et al. (14), we would ex- only four of 18 participants in the present study (average, pect alterations in the twitch amplitude of j1 to +3%. j7%), thus rendering this pathophysiology unlikely in the Hence, we believe that differences in lung volume did not present study. Peribronchial edema, on the other hand, may affect the outcome of the present study.

RESPIRATORY MUSCLE FUNCTION IN ULTRAMARATHON Medicine & Science in Sports & Exercised 525

Copyright © 2015 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited. Moreover, circadian effects could theoretically have biased mostly results from peripheral changes while central mecha- our findings because Fregonezi et al. (10) have shown dif- nisms seem to be of much less importance. The latter is in ferences of up to 3% for MIP and 10% for MEP between contrast to existing data on locomotor muscle fatigue and morning and afternoon measurements. After-race measure- suggests distinctly different profiles of neuromuscular fa- ments of the present study were performed as soon as the race tigue between locomotor and respiratory muscles after an was finished, irrespective of the time of the day relative to the ultraendurance event. Further investigations should be per- baseline measurements. However, after-race reductions in formed to clarify the role of peripheral and central mecha- respiratory muscle strength here clearly outmatch reported nisms in the development of inspiratory and expiratory diurnal variations. Theoretically, sleep deprivation could also muscle fatigue and their potential link with altered postural contribute to reduction in respiratory muscle strength, but control and performance during these types of competition. observations after a 20-h race do not speak in favor of such an effect (34). In addition, for lung function parameters, potential confounding effects of measurements at different times in the We thank all the subjects for their time and maximal efforts put into this study. We sincerely acknowledge the organizers and med- circadian rhythm and the presence of sleep deprivation should ical commission of the North FaceÒ Ultra-Trail du Mont-BlancÒ 2012 be considered as potential confounders. However, as for re- and also the assistance of Dr. Roger Ouillon and Dr. Pascal Edouard spiratory muscle strength, the differences in lung function for conducting medical inclusions and Philippe Gimenez and Marle` ne Giandolini for conducting maximal incremental tests. variables exceeded the reported effects of the circadian Moreover, we would like to thank Dr. David Walker and Dr. Hans- rhythm and sleep deprivation (33). Joachim Kabitz for their technical support regarding the setup for measuring mouth twitch pressure. This study was not funded. CONCLUSIONS Samuel Verges and Christina M. Spengler contributed equally to this study. From the present findings, we conclude that 1) both in- The authors declare that no conflict of interest exists regarding this study. spiratory and expiratory muscle fatigue occur after a moun- The results of the present study do not constitute endorsement by tain ultramarathon and that 2) inspiratory muscle fatigue the American College of Sports Medicine. BASIC SCIENCES REFERENCES 1. Allen DG, Lamb GD, Westerblad H. Skeletal muscle fatigue: 12. Gandevia SC, Butler JE, Hodges PW, Taylor JL. Balancing acts: cellular mechanisms. Physiol Rev. 2008;88(1):287–332. respiratory sensations, motor control and human posture. Clin Exp 2. American Thoracic Society/European Respiratory Society. State- Pharmacol Physiol. 2002;29(1–2):118–21. ment on respiratory muscle testing. Am J Respir Crit Care Med. 13. Gorski J, Namiot Z, Giedrojc J. Effect of exercise on metabolism 2002;166(4):518–624. of glycogen and triglycerides in the respiratory muscles. Pflugers 3. Bellemare F, Bigland-Ritchie B. Assessment of human diaphragm Arch. 1978;377(3):251–4. strength and activation using phrenic nerve stimulation. Respir 14. Hamnegård CH, Wragg S, Mills G, et al. The effect of lung Physiol. 1984;58(3):263–77. volume on transdiaphragmatic pressure. EurRespirJ. 1995; 4. Bellemare F, Bigland-Ritchie B. Central components of dia- 8(9):1532–6. phragmatic fatigue assessed by phrenic nerve stimulation. J Appl 15. Hoffman MD, Ong JC, Wang G. Historical analysis of participa- Physiol (1985). 1987;62(3):1307–16. tion in 161 km ultramarathons in North America. Int J Hist Sport. 5. Berry MJ, Weyrich AS, Robergs RA, Krause KM, Ingalls CP. 2010;27(11):1877–91. Ratings of perceived exertion in individuals with varying fitness 16. Janssens L, Brumagne S, Polspoel K, Troosters T, McConnell A. levels during walking and running. Eur J Appl Physiol Occup The effect of inspiratory muscles fatigue on postural control in Physiol. 1989;58(5):494–9. people with and without recurrent low back pain. Spine (Phila Pa 6. Chevrolet JC, Tschopp JM, Blanc Y, Rochat T, Junod AF. Alter- 1976). 2010;35(10):1088–94. ations in inspiratory and leg muscle force and recovery pattern 17. Johnson BD, Babcock MA, Suman OE, Dempsey JA. Exercise- after a marathon. Med Sci Sports Exerc. 1993;25(4):501–7. induced diaphragmatic fatigue in healthy humans. J Physiol. 1993; 7. Crapo RO, Casaburi R, Coates AL, et al. Guidelines for 460:385–405. methacholine and exercise challenge testing-1999. This official 18. Kabitz HJ, Walker D, Walterspacher S, Windisch W. Controlled statement of the American Thoracic Society was adopted by the twitch mouth pressure reliably predicts twitch esophageal pressure. ATS Board of Directors, July 1999. Am J Respir Crit Care Med. Respir Physiol Neurobiol. 2007;156(3):276–82. 2000;161(1):309–29. 19. Ker JA, Schultz CM. Respiratory muscle fatigue after an ultra- 8. de Bruin PF, Watson RA, Khalil N, Pride NB. Use of mouth marathon measured as inspiratory task failure. Int J Sports Med. pressure twitches induced by cervical magnetic stimulation to as- 1996;17(7):493–6. sess voluntary activation of the diaphragm. Eur Respir J. 1998; 20. Kufel TJ, Pineda LA, Mador MJ. Comparison of potentiated and 12(3):672–8. unpotentiated twitches as an index of muscle fatigue. Muscle 9. Enoka RM, Stuart DG. Neurobiology of muscle fatigue. J Appl Nerve. 2002;25(3):438–44. Physiol (1985). 1992;72(5):1631–48. 21. Loke J, Mahler DA, Virgulto JA. Respiratory muscle fatigue after 10. Fregonezi G, Resqueti VR, Cury JL, Paulin E, Brunetto AF. Di- marathon running. J Appl Physiol. 1982;52(4):821–4. urnal variations in the parameters of pulmonary function and re- 22. Mahler DA, Loke J. Pulmonary dysfunction in ultramarathon spiratory muscle strength in patients with COPD [in English, runners. Yale J Biol Med. 1981;54(4):243–8. Portuguese]. J Bras Pneumol. 2012;38(2):257–63. 23. McConnell AK, Caine MP, Sharpe GR. Inspiratory muscle fatigue 11. Gandevia SC. Spinal and supraspinal factors in human muscle following running to volitional fatigue: the influence of baseline fatigue. Physiol Rev. 2001;81(4):1725–89. strength. Int J Sports Med. 1997;18(3):169–73.

526 Official Journal of the American College of Sports Medicine http://www.acsm-msse.org

Copyright © 2015 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited. 24. Miles DS, Doerr CE, Schonfeld SA, Sinks DE, Gotshall RW. 33. Spengler CM, Shea SA. Endogenous circadian rhythm of pulmo- Changes in pulmonary diffusing capacity and closing volume after nary function in healthy humans. Am J Respir Crit Care Med. running a marathon. Respir Physiol. 1983;52(3):349–59. 2000;162(3 Pt 1):1038–46. 25. Millet GP, Millet GY. Ultramarathon is an outstanding model for 34. Temesi J, Arnal PJ, Davranche K, et al. Does central fatigue ex- the study of adaptive responses to extreme load and stress. BMC plain reduced cycling after complete sleep deprivation? Med Sci Med. 2012;10:77. Sports Exerc. 2013;45(12):2243–53. 26. Millet GY, Lepers R. Alterations of neuromuscular function after 35. Temesi J, Rupp T, Martin V, et al. Central fatigue assessed by prolonged running, cycling and skiing exercises. Sports Med. transcranial magnetic stimulation in ultratrail running. Med Sci 2004;34(2):105–16. Sports Exerc. 2014;46(6):1166–75. 27. Millet GY, Tomazin K, Verges S, et al. Neuromuscular conse- 36. Verges S, Bachasson D, Wuyam B. Effect of acute hypoxia on quences of an extreme mountain ultra-marathon. PLoS One. 2011; respiratory muscle fatigue in healthy humans. Respir Res. 2010; 6(2):e17059. 11:109. 28. National Heart, Lung, and Blood Institute. NHLBI workshop sum- 37. Verges S, Notter D, Spengler CM. Influence of diaphragm and rib mary. Respiratory muscle fatigue. Report of the Respiratory Muscle cage muscle fatigue on breathing during endurance exercise. Fatigue Workshop Group. Am Rev Respir Dis. 1990;142(2):474–80. Respir Physiol Neurobiol. 2006;154(3):431–42. 29. Ortenblad N, Westerblad H, Nielsen J. Muscle glycogen stores and 38. Verges S, Schulz C, Perret C, Spengler CM. Impaired abdominal fatigue. J Physiol. 2013;591(18):4405–13. muscle contractility after high-intensity exhaustive exercise assessed 30. Quanjer PH, Stanojevic S, Cole TJ, et al. Multi-ethnic reference by magnetic stimulation. Muscle Nerve. 2006;34(4):423–30. values for spirometry for the 3-95-yr age range: the global lung 39. Warren GL, Cureton KJ, Sparling PB. Does lung function limit function 2012 equations. Eur Respir J. 2012;40(6):1324–43. performance in a 24-hour ultramarathon? Respir Physiol. 1989;

31. Ross E, Middleton N, Shave R, George K, McConnell A. Changes 78(2):253–63. SCIENCES BASIC in respiratory muscle and lung function following marathon run- 40. Wilcox PG, Wakai Y, Walley KR, Cooper DJ, Road J. Tumor ning in man. J Sports Sci. 2008;26(12):1295–301. necrosis factor alpha decreases in vivo diaphragm contractility in 32. Similowski T, Straus C, Attali V, Duguet A, Derenne JP. Cervical dogs. Am J Respir Crit Care Med. 1994;150(5 Pt 1):1368–73. magnetic stimulation as a method to discriminate between dia- 41. Wilson SH, Cooke NT, Edwards RH, Spiro SG. Predicted normal phragm and rib cage muscle fatigue. J Appl Physiol (1985). 1998; values for maximal respiratory pressures in caucasian adults and 84(5):1692–700. children. Thorax. 1984;39(7):535–8.

RESPIRATORY MUSCLE FUNCTION IN ULTRAMARATHON Medicine & Science in Sports & Exercised 527